Content-based Audio Classification and Retrieval using SVM Learning
- Guodong Guo ,
- Stan. Z. Li
Published by Institute of Electrical and Electronics Engineers, Inc.
In this paper, a support vector machines (SVMs) based method is proposed for content-based audio classification and retrieval. Given a feature set, which in this work is composed of perceptual and cepstral feature, optimal class boundaries between classes are learned from training data by using SVMs. Matches are ranked by using distances from boundaries. Experiments are presented to compare various classification methods and feature sets.
© 2000 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.